Open Arabic LLM - Arabic MMLU General Knowledge (Middle School) — leaderboard
Metric: Accuracy (%). Source: huggingface.co. 163 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Ultiima-72B | 83.72 |
| 2 | Rombos-LLM-V2.5-Qwen-72B | 82.56 |
| 3 | calme-2.1-qwen2.5-72B | 81.98 |
| 4 | Qwen 2.5 72B Instruct | 81.4 |
| 5 | calme-2.2-qwen2.5-72B | 79.65 |
| 6 | gemma-3-27B-pt | 77.91 |
| 7 | SILMA-9B-Instruct-v1.0 | 77.91 |
| 8 | Qwen 2.5 32B Instruct | 76.74 |
| 9 | Qwen2.5-32B-Instruct-abliterated-v2 | 76.74 |
| 10 | Awqward2.5-32B-Instruct | 76.74 |
| 11 | Qwentile2.5-32B-Instruct | 76.74 |
| 12 | Saka-14B | 76.16 |
| 13 | Qwen2.5-32B-Instruct-CFT | 76.16 |
| 14 | Qwen 2.5 72B | 75.58 |
| 15 | lambda-qwen2.5-32B-dpo-test | 75.58 |
Interactive version: theaggregate.ai/benchmark?slug=open-arabic-llm-arabic-mmlu-general-knowledge-middle-school · How the rankings work · Data refreshed daily, snapshot 2026-07-22.